So-called virtual, that is gestural and/or tactile and vocal man-machine interfaces, as opposed to interfaces where the user acts on mechanical sensors such as keyboard keys or switches, are most often 3D cameras in the field of gestural interface and active surfaces, often transparent and integrated on the display screens, in the field of tactile interface. These tactile interfaces are widespread under the name of “touchpad” for multiple industrial and consumer applications, such as for example smartphones, home automation or game interfaces.
The screen enables to display for example a keyboard or the software interface the image of which varies in response to actions from the user, which generally consist of a displacement, a tapping or even a prolonged contact of the finger or a stylus onto the screen surface in order to perform controls. More recently, so-called gestural interfaces have developed to meet the increasingly great complexity of digital products emanating from the convergence of communication and entertainment information technologies. These interfaces are most often based on a 3D camera and an image processing capable of interpreting at a long distance, up to 5 meters, the movements of the body or the hand to interact with the screen.
Among virtual controls, can also be mentioned the vocal control which enables the request to be performed from the voice.
Among existing tactile technologies, capacitive technologies are frequently used because:                they do not require exerting any mechanical action onto the screen surface unlike resistive techniques for example,        they are well-adapted to the meshing of the screen surface by a sensor network directly integrated to the surface thereof, which enables a much more compact and robust integration than with optical techniques for example which require a network of transmitters-receivers raised with respect to the detection surface.        
It is more and more attempted in virtual interfaces to have a proximity detection function, which enables new man-machine interaction modes to be created without any contact. Sensors must then be able to detect displacements or shapes within several centimeters, in a sufficiently accurate and determined manner to be able to transform them into controls.
If optical technologies cannot be dispensed with when accurately detecting movements at a very long distance (beyond 40 centimeters), capacitive technologies turn out to be very well adapted to interfaces without proximity contact, in particular because all sensors can be integrated to a non-planar active surface, for example a screen, without dead area or without requiring external devices (cameras).
Amongst capacitive tactile interfaces, the most used techniques are generally based on the charge transfer principle. They enable sensitivity in the order of one picofarad to be obtained but they are not adapted to make a gestural interface. Indeed, their sensitivity does not enable the approach of a finger to be detected because the capacity generated does not then exceed a few hundreds picofarads. Stray capacitances present in this type of sensor and in electronics prevent any sensitivity improvement.
Moreover, these techniques are very sensitive to electromagnetic disturbances, stray electrical charges and electric conduction of the dielectric covering the electrodes. When the air relative humidity is higher than 60%, most dielectrics become slightly electrically conducting and charges generated by the electronics are modified, which disturbs the capacitive measurement.
Among technical solutions employed to make non-tactile capacitive interfaces, some are known which implement a pseudo-guard enabling the stray capacitances to be strongly reduced in the sensor and electronics. However, these techniques only enable at most one order of magnitude on sensitivity to be gained, which enables the presence of a finger to be detected at only a few mm on the sensitive surface of the sensor.
Regarding capacitive gestural interfaces, it is known from document U.S. Pat. No. 6,847,354 to Vranish, capacitive sensors implemented with an active guard. This active guard is created using a unity gain amplifier which generates in the guard a voltage with an amplitude identical to that of the measurement electrode.
The drawback of this method is that the electronics generates by a physical principle stray capacitances which correspond to the sum of the input capacitances of the amplifier to generate the active guard, also called pseudo-guard, and of the circuit to excite the measurement electrode. These stray capacitances easily reach one picofarad and add up to the capacitance to be measured, which only represent one hundredth of this total value.
Moreover, the capacitance measurement is not direct because the Vranish electronics obtains the image of the capacitance to be measured by measuring its current via a reference resistor. This resistor generates a stray phase shift of the electrical current which strongly degrades the detection quality, or of demodulation of the signal representing the capacitance to be measured. Another drawback is the crosstalk level between different electrodes. Indeed, each electrode is excited by an operational amplifier the gain of which is approximately a unity one. Any gain deviation between different amplifiers causes a high further stray capacitance.
These drawbacks do not enable the position of an object such as a finger or a hand to be detected and located at several centimeters or even tens of centimeters with each sensor.
On the other hand, capacitive technologies implemented for making gestural interfaces have been developed most often in an attempt to be integrated to screens or at least substantially planar surfaces. The structures of sensors are arrays, as in U.S. Pat. No. 6,847,354, and interfaced with electrode structures disposed as a grid X-Y. These technologies are poorly adapted to the instrumentation of surfaces with more complex shapes.
In particular, they are hardly applicable to devices for virtual controls or gestural interfaces based not on an instrumented planar surface but on other types of geometries including cavities, reliefs, undulations simulating for example keys, buttons or wrapping the user wherein sensors can be disposed as various geometries, sometimes under dielectric materials with a high thickness, and must sometimes be able to be disposed and managed independently of one another.
Document FR2884349 to Rozière is known, which discloses a system based on capacitive sensors including a single measurement electrode. The device enables objects to be detected beyond 10 cm with each electrode, thanks to a floating bridge electronics. The measurements from the electrodes are sequentially read by means of a scrutinizing system.
However, this device, intended to equip walls of movable instruments, is specifically designed to cover anti-collision applications, and does not cover the gestural interface applications.
Whereas gestural interfaces will have an increasing importance, the touch however remains prevailing, even in virtual interfaces capable of detecting movements at a distance, for psychological and security reasons in particular. Indeed, on the one hand the tactile feeling promotes comfort to the user, on the other hand the touch allows efficient validation of commands in operating contexts where security is important (medical equipment, vehicles).
The object of the present invention is to provide a gestural interface and a gestural and tactile interface, comprised of a device and a method for monitoring actions and inputting controls, which is compact and integrable in a variety of environments, while allowing:                anticipating actions by accurate measurements of displacements without contact, and        securely validating commands through detection of physical actions on a surface.        